Optimizing Electric Vehicle Charging Infrastructure through Machine Learning: A Study of Charging Patterns and Energy Consumption
DOI: 10.3991/ijim.v18i21.50843
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study addresses the critical challenge of optimizing electric vehicle (EV) charging infrastructure in metropolitan areas, specifically focusing on the inefficient placement of charging stations and inconsistent demand patterns. The rapid adoption of EVs has created a pressing need for efficient infrastructure to support sustainable urban transportation, yet issues such as idle time, grid strain, and poor station distribution hinder market penetration. The authors propose a machine learning-based framework to predict energy consumption and optimize the placement of EV charging stations (EVCSs), aiming to enhance resource allocation and user experience. The research utilizes a comprehensive dataset comprising 148,136 charging transactions from city-owned EV charging stations in Boulder, Colorado. The dataset includes detailed transaction records such as start and end times, duration, energy consumption in kWh, and station locations. The methodology involves several steps: first, exploratory data analysis was conducted to identify busy hours and high-consumption stations. Second, geographic coordinates were generated for station addresses using OpenStreetMap (OSM) to enable spatial analysis. Finally, machine learning models, specifically KNeighborsRegressor and RandomForestRegressor, were trained to predict energy consumption based on geographic and temporal features. The study also incorporates a user-centric recommendation system to suggest charging stations based on remaining energy levels. The analysis revealed significant fluctuations in charging demand, with peak activity occurring during morning commute hours (6 AM–10 AM) and evening hours (4 PM–11 PM). Conversely, early morning and midday periods exhibited lower activity levels. Spatial analysis identified specific high-consumption hubs, such as the “BOULDER/BASELINE ST1” station, which recorded 1,136.754 kWh of power consumption, while other stations showed minimal usage. Charging activity varied by ZIP code, with areas like 80302 showing high demand during morning and early afternoon peaks, whereas 80304 and 80305 demonstrated low, consistent activity. In terms of model performance, the KNeighborsRegressor algorithm demonstrated superior prediction accuracy compared to the RandomForestRegressor for estimating energy consumption. The findings provide actionable insights for infrastructure planning and load management. By identifying peak demand times and high-utilization locations, policymakers and network owners can optimize station placement and resource allocation to mitigate grid strain and reduce user idle time. The proposed machine learning framework supports the development of smart charging networks that respond to real-time demand, ultimately promoting EV adoption and enhancing the efficiency of urban transportation systems. The study highlights the importance of data-driven approaches in addressing the complexities of EV infrastructure expansion.
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-06-20 |
| archive | success | canonical_url | — | — | 1 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.